You can't lose money in stocks if you hang in long enough. Could there be a giant fallacy in this line of reasoning?
History, as written by the winners

By Philip E. Ross

I am a bull

IT HAS COME to be taken as gospel: Stocks are a can't-lose investment. They have bad years, but they win out in the end. If you are young and can afford to sit through the inevitable corrections, you ought to have all your savings in stocks.

Need some money in the interim? You just borrow against stocks. Hang in for at least 30 years and you can't lose money.

How do we know this? From the statistics. In every 30-year period since 1871, stocks have handily beaten every other financial asset, including cash. Since 1926, stocks, with dividends included, have made money in every 20-year period.

The gospel is preached these days by just about every financial planner, financial columnist and stockbroker you run into. The problem with it is that it may be based on a faulty interpretation of history.

Listen to Paul Samuelson, the distinguished Nobel-winning economist and also an astute investor: "Most people have been encouraged to become long-term stock investors by a line of reasoning that is incorrect," he says. "That doesn't mean I can tell someone not to be a long-term investor in equities, only that the current bull market is being fed by a misunderstanding of the law of large numbers."

The misunderstanding he is referring to has to do with survivorship bias. When the pundits say stocks always win, they turn for proof not just to any nation with a history of stock returns but to a particular one: the U.S. The record is long and wonderfully positive. If you had held U.S. stocks since 1816, you would have averaged a terrific 6.7% compound return, above and beyond inflation. Considering that money doubles every 12 years at 6%, you'd need a lot of zeroes to show how much you would have if you had invested $100 in stocks in 1816. You'd need a computer to tell you.

But is this country typical? In the early 19th century the U.S. was just beginning a two-century-long expansion into a resource-rich continent, unpunctuated by a revolution or invasion. Not coincidentally, long continuous records are lacking for countries that have had periods of acute discomfort for owners of equity capital—countries like Germany, Russia, Poland and Japan. Probably there weren't any stocks in Russia in the very early 19th century, but if there had been, we're not at all sure they would have been a great investment. Maybe stocks are not the sure winners they are cracked up to be.

The New York Stock Exchange has maintained continuous trading since its inception in 1792. Tack on the London exchange, and the unbroken ticker tape stretches all the way back to 1773. More importantly, the two big Anglo-Saxon exchanges kept going through the 1930s and 1940s, when their less fortunate cousins either suffered big breaks in trading or simply winked out.

Financial historians depend on Anglo-Saxon data because they need long data series to draw conclusions about long-term risk. The trouble is they have picked two big winners. Great Britain was the preeminent economic power in the 19th century and into the 20th; the U.S. has been for most of the 20th. How good, though, is triumphalist history, written from the winner's point of view?

"Economists commonly take historical data from the winning countries and extrapolate that into the future," warns William Goetzmann, a professor of finance at the Yale School of Management. "If you take data from the losing countries, you get a different picture."

Since 1993 Goetzmann and Philippe Jorion, of the University of California at Irvine, have been doing just that, collecting data from countries where economists have lacked stock market data because of breaks in trading.  

Honda   What happens when stock trading stops? You could suffer a temporary inconvenience, as Swiss shareholders did when their exchange closed during a general mobilization in 1940 against a feared German invasion. Or you could lose everything, as did those who invested in the St. Petersburg exchange before the Russian Revolution of 1917. The chart, Survivors rule, gives the benefit of the doubt to markets like Tokyo's where there was a prolonged disruption: It simply ignores missing years. But Goetzmann speculates that half of equity value vanishes in such years.

So what are some of the really bad things that might happen to our stock market? Take a look at another seemingly secure country: Britain in the 1970s.

"In the early 1970s the U.K. went through an incredible period of turmoil," says Goetzmann. "When we looked at the pages of The Economist, we found articles questioning whether capitalism would survive in Britain at all.

"In a slightly different world the U.S. exchanges might have met their end in the 1930s," he goes on. "Had the Senate investigation into the stock market crash discovered one more scandal, it might have tipped the balance in favor of those who wanted to end equities trading for good. It seems hard to imagine, but then, no one who comes off a long period of stable finance is likely to see that kind of disaster looming until it's too late.

"After the Russian Revolution, the British press was really skeptical that the mining shares on the St. Petersburg exchange would be really worthless," says Goetzmann. "They figured they [the mines] were too productive for the Bolsheviks to be so foolish as to kick out management. So the Bolsheviks took over the mines, and then they said, 'Okay, but they'll still need foreign capital.' Then the foreign capital was expropriated. There was denial all down the way, until the shares were really worthless."

Goetzmann doesn't disagree that most stock markets have done well most of the time. He simply finds enough disturbing exceptions to the rule to give him pause. Looking only at price returns, since dividend data is extremely scanty in most of the world, he and Jorion plotted average annual inflation-adjusted returns for 39 countries. A sampling is shown in the chart on page 206.

For a number of countries, including Belgium, Spain, New Zealand and Italy, the number is negative. In this context the U.S. stock market is a bit of a freak, combining a long uninterrupted data series with a very positive return.

What would you get with dividends added back in? No one knows for a lot of these countries, but dividends would probably be high enough to put just about all of the 39 into the positive column for real total return. In Spain, though, the return would be pretty meager—in the vicinity of 1% or 2% over stretches covering at least half a century.

Japan is an interesting case. During the period 1921-44, the professors calculate, stock prices in Tokyo did not keep up with inflation. During 1949-95 they raced ahead. What happened to stockholders between 1944 and 1949?  

Survivors rule

Stocks beat bonds

  Good question. We know that some controlling owners of companies—for example, Kononsuke Matsushita of Matsushita Electric Industrial Co.—were able to retain their property after the war. But just what happened to public holders of shares traded on the Tokyo exchange remains a mystery.

Those who bought Japanese equities in late 1989, when the Nikkei was at 39,000, have lost nearly half their money after eight years. It's impossible to lose money in stocks if you hold them for 30 years? Maybe, maybe not. Remember this: From 1966 to 1982, the U.S. stock market, as measured by the Dow industrials, went nowhere. It has since more than made up for that, but this didn't help someone who bought in 1966 and died in 1976.

It has long been a mystery among academics why stocks have done so much better than bonds throughout U.S. history. Goetzmann and Jorion think they have part of the answer: It's a rational response on investors' part to the risk that something catastrophic will happen someday to the New York Stock Exchange.

Not everyone agrees with them. Jeremy Siegel, a professor at Wharton and an authority on financial history, thinks that stocks have done well because savers have been too foolish to want them. Had people been rational, that is, they would have accepted smaller returns on stocks and insisted on better returns on bonds, pushing the two asset classes closer to equality (see chart, Stocks beat bonds).

"People have an irrational fear of short-term risk," Siegel says. If he's right, and if you have no such fears, you can improve your lot by dumping your Treasurys and buying stocks. He denies that investor jitters reflect a rational fear that what happened to foreign exchanges might one day happen to ours. Reason: Bonds are equally vulnerable to political disaster.

Paul Samuelson is not a bear, but he agrees with Professor Goetzmann that people who are not 100% in stocks are not necessarily stupid.

"There is a survival bias," Samuelson says. "My hunch is that if you were able to eradicate all of that bias, you would remove part of the demonstrated superiority of equities in the last 150 years over alternative investments, such as bonds, money market funds and bank accounts.

"It is in terms of that rational calculation that you should determine whether you are a 10% equities investor, or 50% or 100%," says Samuelson.

He adds, a bit sadly: "I have students of mine—Ph.D.s—going around the country telling people it's a sure thing to be 100% invested in equities, if only you will sit out the temporary declines. It makes me cringe."  


I am a bull

ROGER IBBOTSON, Yale economist and vendor of financial databases, has almost all of his liquid investments in common stocks. He thinks stocks always have been, and always will be, much better than bonds in long-term performance. He has a lot of company (see story).

What sets Ibbotson apart is the passion with which he preaches the equities gospel. His view: You ain't seen nothing yet. By year-end 2001 the Dow Jones industrial average will hit the magic 10,000. By 2015 it will clear 34,000. In 2020 look for the average to hit 53,000.

Let's see. I put $100,000 in stocks today, I'll have $757,000 in 23 years without even counting dividends.

Ibbotson will turn 77 that year. If he's right, he's going to be one rich professor. He's already doing quite well, as principal owner of Chicago-based Ibbotson Associates, which sells software and data that money managers use for statistical analysis.

What makes him so confident? Not fundamentals. Stocks, after all, are trading at historically steep multiples of their earnings, dividends and book value. No, his market calls are predicated on just one thing: past performance. What was, will be.

Along the way, there are corrections, of course, but intrepid investors just ride them out. With one exception, market dips have never lasted more than two years. That exception was a doozy, a 90% decline in stock prices between 1929 and 1932. So the professor doesn't say that prolonged crashes can't happen, only that they are unusual. The corrections of 1987 and 1990 were much shorter and are mere blips in a long-term chart. "Over the past 71 years, stock prices have moved up in two of every three years," says Ibbotson.

He goes on: "There's been no 20-year period where you lost money in the stock market, and only two periods of ten years, 1929-38 and 1930-39, when stocks lost ground." He is putting stocks in their best light: including dividends but not adjusting for inflation.

What if you look at price changes only and adjust for inflation (as in the international scatter chart, Survivors rule)? Then you see more bearish stretches, including, in the U.S., the decade of 1965-74.

Given that recent performance has been well ahead of the long-run average, don't stocks have to take a bit of a breather?

Nonsense, says Ibbotson: "There's no evidence to support the idea that after a period of high returns we get a period of low returns. We don't have to offset the high returns. It's possible we'll mark time. It's possible, but not likely."

So why isn't everyone 100% invested in stocks? "There's a one-in-three chance that stocks will go down in any year. Most people aren't willing to take that risk. Risk causes prices to be lower and makes the returns higher," he maintains. Ibbotson makes it sound too easy.
—Robert Lenzner   


  Survivors rule Older exchanges tend to do best. That suggest that returns to equity are biased by simple survivorship. This chart, covering the years since the 1920s, somwhatn overstates the performence of exchanges with trading breaks.



  Stocks beat bonds

If you believe the past predicts the future, then you will take Professor SiegelĘs advice and put all your long-term savings in the stock market.

Source: Stocks for the long run. Jeremy Segal

June 16, 1997 Issue     © Forbes Inc. 1997

Click to see more like this:

By Philip E. Ross     On the Cover     Investment Guide     From June 16, 1997 Issue

Back to Top    Forbes Today